Business Case

Road safety rules and regulations are designed to prevent the citizens from fatal incidents. Although policies are in place, we observe negligent behaviour of the drivers which lead to serious injuries or death crashes. It is of utmost interest of the authorities to understand and analyse human behaviour to take necessary corrective and preventive actions.

From the given data (i.e the vehicle details, the weather details and the vehicle travelling details) we need to predict Driving Style of the particular Driver whether it is "Agressive", "Normal" or "Vague

In ths particular problem we are intersted in finding the Aggressive Drivers as It may be benificial for stakeholders. The stakeholders are :

  1. The Citizens
  2. The Road Transport authorities
  3. Insurers

  4. Research/ Data Service Providers

    • The Citiens will be benefitted as it will be helpful in reducing the no. of accidents
    • The Road Tansport Authorities the maintainance cost is reduced as we can reduce the number of accidents. Moreover they get insights to design benchmark for qualifications and driver licensing regulations, etc.
    • The Insurers will benifit in analysing underwriting risks, prevent frauds and designing No-claim-discount systems (NCD systems),

It is very important for the business that there can be false positive but if we predict false negative (i.e we predict Aggressive Drivers as Non Aggressive) then it will impact the Stakeholders, thus we have to improve the Recall value

ML Problem Statement

From the given data (i.e the vehicle details, the weather details and the vehicle travelling details) we need to predict Driving Style of the particular Driver whether it is "Agressive", "Normal" or "Vague. These patterns are pre identified in the datset given and one has to infer the style for new datasets by prediction.

As we have understood from the business that, the model built has to be tuned for Recall as False -ves are a strict no. Also have to provide top 10 patterns to identify Aggresive driving Style.

Importing Necessary libraries

In [1]:
# Importing Necessary libraries 
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import seaborn as sns
import plotly
import plotly.offline as pyoff
import plotly.figure_factory as ff
from plotly.offline import init_notebook_mode, iplot, plot
import plotly.graph_objs as go
%matplotlib inline

#Importing lib for Modelbuilding
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
from sklearn.metrics import recall_score
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import GradientBoostingClassifier
from xgboost import XGBClassifier
from sklearn.model_selection import StratifiedKFold
from sklearn.model_selection import cross_val_score
from sklearn.ensemble import RandomForestClassifier
from sklearn.neural_network import MLPClassifier

For getting the output of plot just after cell

In [2]:
init_notebook_mode(connected=True)

Reading the csv files

In [3]:
train_data = pd.read_csv('Train.csv')
In [4]:
train_vehicletravelling_data = pd.read_csv('Train_Vehicletravellingdata.csv')
In [5]:
train_weather_data =pd.read_csv('Train_WeatherData.csv')

Renaming the columns as per Attribute Description

In [6]:
train_data.columns = ['ID','Vehicle_len','Vehicle_Wt','Num_Axle','DrivingStyle']
In [7]:
train_vehicletravelling_data.columns = ['ID','date_time_trip','road_lane','vehicle_speed','pre_vehicle_ID','pre_vehicle_speed','pre_vehicle_wt','pre_vehicle_len','time_gap_pre_vehicle','road_condition']
In [8]:
train_weather_data.columns = ['ID','date_time_trip','air_temp','precipitation_type','precipitation_intensity','rel_humidity','wind_direction','wind_speed','lighting_condition']
In [9]:
print(train_data.shape)
print(train_weather_data.shape)
print(train_vehicletravelling_data.shape)

(12994, 5)
(162566, 9)
(162566, 10)

Merging Data

  1. Merging vehicle travelling data and weather data on 'ID' and 'date and time' column
  2. Merging the above data with train i.e vehicle data on 'ID' column
In [10]:
train_vehicletravelling_weather = pd.merge(train_vehicletravelling_data, train_weather_data, how='inner', on=['ID','date_time_trip'])
In [11]:
train = pd.merge(train_vehicletravelling_weather,train_data, how='outer', on=['ID'])
In [12]:
train.columns
Out[12]:
Index(['ID', 'date_time_trip', 'road_lane', 'vehicle_speed', 'pre_vehicle_ID',
       'pre_vehicle_speed', 'pre_vehicle_wt', 'pre_vehicle_len',
       'time_gap_pre_vehicle', 'road_condition', 'air_temp',
       'precipitation_type', 'precipitation_intensity', 'rel_humidity',
       'wind_direction', 'wind_speed', 'lighting_condition', 'Vehicle_len',
       'Vehicle_Wt', 'Num_Axle', 'DrivingStyle'],
      dtype='object')
In [13]:
train.shape
Out[13]:
(162566, 21)
In [14]:
train_driving_style = train.DrivingStyle.replace({1: 'Aggressive', 2: 'Normal', 3: 'Vague'})

Exploratory Data Analysis

In [129]:
train.head()
Out[129]:
ID road_lane vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle road_condition air_temp precipitation_type precipitation_intensity rel_humidity wind_direction wind_speed lighting_condition Vehicle_len Vehicle_Wt Num_Axle DrivingStyle
0 DR_24526 1 81 87 16986 941 94.0 Dry 7.0 clear None 95.0 146.0 1.0 daylight 1265 10243 2 2
1 DR_24526 1 88 81 1708 551 11.0 Dry 7.0 clear None 95.0 124.0 0.0 daylight 1265 10243 2 2
2 DR_24526 1 88 88 22892 1698 4.0 Dry 7.0 clear None 95.0 124.0 0.0 daylight 1265 10243 2 2
3 DR_24526 2 84 89 1945 544 127.0 Dry 7.0 clear None 95.0 124.0 0.0 daylight 1265 10243 2 2
4 DR_24526 1 89 88 13787 1893 42.0 Dry 7.0 clear None 95.0 124.0 0.0 daylight 1265 10243 2 2
In [128]:
train.describe().transpose()
Out[128]:
count mean std min 25% 50% 75% max
vehicle_speed 162566.0 83.455483 9.375122 8.0 78.0 83.0 88.0 161.0
pre_vehicle_speed 162566.0 83.458817 9.373120 0.0 78.0 83.0 88.0 161.0
pre_vehicle_wt 162566.0 5017.562793 7399.315434 3.0 1502.0 1862.0 2669.0 69548.0
pre_vehicle_len 162566.0 790.775285 481.944297 102.0 527.0 560.0 701.0 2981.0
time_gap_pre_vehicle 162566.0 105.340358 175.859156 1.0 7.0 45.0 123.0 1797.0
air_temp 162566.0 4.654237 3.206222 -13.0 2.0 5.0 7.0 24.0
rel_humidity 162566.0 60.647509 18.154208 16.0 46.0 58.0 76.0 97.0
wind_direction 162566.0 182.461178 88.347763 6.0 152.0 180.0 208.0 360.0
wind_speed 162566.0 4.188648 3.004563 0.0 1.0 4.0 7.0 17.0
Vehicle_len 162566.0 865.518349 495.155881 155.0 550.0 577.0 1060.0 2337.0
Vehicle_Wt 162566.0 6020.270481 7972.925231 44.0 1625.0 2013.0 6220.0 57230.0
In [158]:
train.shape
Out[158]:
(162566, 19)

Observation :

  1. There are 19 identical columns and 162566 rows after merging the data
  2. There ar two options either we can work with 162566 records or we can converge the data according to vehicle details table and converging the number of records to 12994

Tried both approaches attaching both files

Checking for Null Values

In [15]:
train.isnull().sum()
Out[15]:
ID                            0
date_time_trip                0
road_lane                     0
vehicle_speed                 0
pre_vehicle_ID                0
pre_vehicle_speed             0
pre_vehicle_wt                0
pre_vehicle_len               0
time_gap_pre_vehicle       2455
road_condition                0
air_temp                   2057
precipitation_type            0
precipitation_intensity       0
rel_humidity               2105
wind_direction             2114
wind_speed                 2464
lighting_condition            0
Vehicle_len                   0
Vehicle_Wt                    0
Num_Axle                      0
DrivingStyle                  0
dtype: int64
In [16]:
plt.figure(figsize=(10,10))
train.isnull().sum().plot(kind='bar',
                                  figsize=(10,6),
                                  color="blue",
                                  alpha = 0.7,
                                  fontsize=13)
plt.title("NA's (in numbers)")
plt.show()

Correlation Between Variables

In [17]:
target = train["DrivingStyle"]
train_without_target = train.drop(["DrivingStyle"],axis=1)
corr = train_without_target.corr()
f, ax = plt.subplots(figsize=(10, 6))
sns.heatmap(corr,cmap='coolwarm',vmax=.3)
Out[17]:
<matplotlib.axes._subplots.AxesSubplot at 0xf590278>

Observation:

  1. preceeding vehicle length and preceeding vehicle weight are highly correlated
  2. vehicle length and vehicle weight and Number of Axles are positively correlated
  3. Wind Speed and Relative Humidity Negatively Correlated
  4. Road Lane and Vehicle Speed are correlated
  5. Road Lane and Preceeding Vehicle speed is correlated
  6. Vehicle Length and time gap between preceeding vehicle positively correlated

Analysing Target Varible (Driving Style)

In [18]:
# plt.figure(figsize=(10,10)
train.DrivingStyle.value_counts().plot(kind='bar',
                                 figsize=(10,6),
                                  color="red",
                                  alpha = 0.7,
                                  fontsize=13)
plt.title('Driving Style (in numbers)')
plt.xlabel('Driving Style')
plt.ylabel('Count')
plt.show()

Plotting Bar Graph For Categorical Variables

In [19]:
def generate_layout_bar(col_name):
    layout_bar = go.Layout(
        autosize=False, # auto size the graph? use False if you are specifying the height and width
        width=800, # height of the figure in pixels
        height=600, # height of the figure in pixels
        title = "Distribution of {} column".format(col_name), # title of the figure
        # more granular control on the title font 
        titlefont=dict( 
            family='Courier New, monospace', # font family
            size=14, # size of the font
            color='black' # color of the font
        ),
        # granular control on the axes objects 
        xaxis=dict( 
        tickfont=dict(
            family='Courier New, monospace', # font family
            size=14, # size of ticks displayed on the x axis
            color='black'  # color of the font
            )
        ),
        yaxis=dict(
#         range=[0,100],
            title='Percentage',
            titlefont=dict(
                size=14,
                color='black'
            ),
        tickfont=dict(
            family='Courier New, monospace', 
            size=14,
            color='black'
            )
        ),
        font = dict(
            family='Courier New, monospace',
            color = "white",
            size = 12
                )  
        )
    return layout_bar
In [20]:
def plot_bar(col_name):
    # create a table with value counts
    temp = train[col_name].value_counts()
    # creating a Bar chart object of plotly
    data = [go.Bar(
            x=temp.index.astype(str), # x axis values
            y=np.round(temp.values.astype(float)/temp.values.sum(),4)*100, # y axis values
            text = ['{}%'.format(i) for i in np.round(temp.values.astype(float)/temp.values.sum(),4)*100],
            textposition = 'auto', 
        marker = dict(color = '#0047AB'),)] 
     
    layout_bar = generate_layout_bar(col_name=col_name)
    fig = go.Figure(data=data, layout=layout_bar)
    return iplot(fig)
In [21]:
sns.catplot(x="road_lane", kind="count", palette="ch:.25", data=train);

Observation:

Two lanes on the road and the traffic is equal on both the lanes

In [22]:
sns.catplot(x="road_condition", kind="count", palette="ch:.30", data=train);

Observation :

Road Condition has four labels Dry, Wet, Snow Covered and Visible Tracks

In [23]:
sns.catplot(x="lighting_condition", kind="count", palette="ch:.30", data=train);
In [24]:
plot_bar('precipitation_type')

Observation :

Most of the time 93% weather is clear according to the data

In [25]:
plot_bar('precipitation_intensity')

Observation :

  1. There is a blank category
  2. 93.24% of the precipitation_intensity column has None value as from the precipitation type cloumn we see most of the time the weather is clear
In [26]:
#Replacing the blank category to a new Category
train.precipitation_intensity = train.precipitation_intensity.replace(to_replace=' ', value='NotKnown')
In [27]:
 plot_bar('lighting_condition')

Observation :

Lighting Condition is mostly Night time i.e. the data that is collected is mostly during Night Time

In [28]:
plot_bar('Num_Axle')

Observation :

Most of the vehicles have 2 axles

Visualization for continuous variable

In [29]:
def plotHistogramGraph(col_name):
    data = [go.Histogram(x=train[col_name],
       marker=dict(
        color='#CC0E1D',   
    ))]
    layout = go.Layout(title = "Histogram of {}".format(col_name))
    fig = go.Figure(data= data, layout=layout)
    iplot(fig)
In [30]:
plotHistogramGraph('vehicle_speed')

Observation:

Mostly vehicle speed is between 70 to 100

In [31]:
plotHistogramGraph('wind_direction')

Observation:

The Wind direction is mostly between 160 to 190 degrees

In [32]:
train.air_temp.sample(frac =0.1)
plt.figure(figsize=(15,10))
sns.countplot(x='air_temp', data=train)
Out[32]:
<matplotlib.axes._subplots.AxesSubplot at 0x17d7c2e8>

Observation :

The Min temp is -13 degree and Max is 24 degree. which may mean that the data set belongs to colder region

In [33]:
DrivingStyle_dict = {1:'1',
           2:'2',
           3:'3',
           }
train = train.replace({"DrivingStyle":DrivingStyle_dict })

Relation between vehicle speed and lighting Condition

In [34]:
driving_air_temp = train.groupby(['DrivingStyle','air_temp']).size().to_frame()
driving_air_temp = driving_air_temp.reset_index()
driving_air_temp.columns = ['DrivingStyle','air_temp','Count']
In [35]:
np.random.seed(123)
driving_air_temp.sample(frac =0.1)
Out[35]:
DrivingStyle air_temp Count
8 1 -5.0 24
70 3 -11.0 37
82 3 1.0 5991
28 1 15.0 18
63 2 20.0 61
0 1 -13.0 8
5 1 -8.0 14
50 2 7.0 9361
81 3 0.0 325
4 1 -9.0 27
In [36]:
trace1 = go.Bar(x = driving_air_temp.air_temp[driving_air_temp.DrivingStyle=='1'],
               y = driving_air_temp.Count[driving_air_temp.DrivingStyle=='1'],
                text = driving_air_temp.Count[driving_air_temp.DrivingStyle=='1'],
                textposition = 'auto',
               name = 'Aggressive')
trace2 = go.Bar(x = driving_air_temp.air_temp[driving_air_temp.DrivingStyle=='2'],
               y = driving_air_temp.Count[driving_air_temp.DrivingStyle=='2'],
                text = driving_air_temp.Count[driving_air_temp.DrivingStyle=='2'],
                textposition = 'auto',
               name = 'Normal')

trace3 = go.Bar(x = driving_air_temp.air_temp[driving_air_temp.DrivingStyle=='3'],
               y = driving_air_temp.Count[driving_air_temp.DrivingStyle=='3'],
                text = driving_air_temp.Count[driving_air_temp.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Vague')

data = [trace1, trace2, trace3]
layout = go.Layout(width = 1000,
#                    barmode='stack',
                   height = 600,title = 'Air Temperature and Driving Style',
                  xaxis =  dict(title = 'Air Temperature'),
                   yaxis =  dict(title = 'Counts',range=[0, driving_air_temp.Count+10]))
fig = go.Figure(data=data, layout=layout)
iplot(fig)

Relation between Driving Style and Precipitation Type

In [37]:
# train.precipitation_type
driving_prec_veh = train.groupby(['DrivingStyle','precipitation_type']).size().to_frame()
driving_prec_veh = driving_prec_veh.reset_index()
driving_prec_veh.columns = ['DrivingStyle','precipitation_type','Count']
In [38]:
trace1 = go.Bar(x = driving_prec_veh.precipitation_type[driving_prec_veh.DrivingStyle=='1'],
               y = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='1'],
                text = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='1'],
                textposition = 'auto',
               name = 'Aggressive')
trace2 = go.Bar(x = driving_prec_veh.precipitation_type[driving_prec_veh.DrivingStyle=='2'],
               y = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='2'],
                text = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Normal')

trace3 = go.Bar(x = driving_prec_veh.precipitation_type[driving_prec_veh.DrivingStyle=='3'],
               y = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='3'],
                text = driving_prec_veh.Count[driving_prec_veh.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Vague')

data = [trace1, trace2, trace3]
layout = go.Layout(width = 1000,
#                    barmode='stack',
                   height = 600,title = 'Driving Style with Precipitation Type',
                  xaxis =  dict(title = 'Precipitation Type'),
                   yaxis =  dict(title = 'Counts',range=[0, driving_prec_veh.Count.max()+10]))
fig = go.Figure(data=data, layout=layout)
iplot(fig)

Observation:

Some people drive aggressively inspite of rain and snow

In [39]:
Driving_Road_Condition = train.groupby(['DrivingStyle','road_condition']).size().to_frame()
Driving_Road_Condition = Driving_Road_Condition.reset_index()
Driving_Road_Condition.columns = ['DrivingStyle','road_condition','Count']
In [40]:
trace1 = go.Bar(x = Driving_Road_Condition.road_condition[Driving_Road_Condition.DrivingStyle=='1'],
           y = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='1'],
            text = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='1'],
            textposition = 'auto',
           name = 'Aggresive')
trace2 = go.Bar(x = Driving_Road_Condition.road_condition[Driving_Road_Condition.DrivingStyle=='2'],
               y = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='2'],
                text = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='2'],
                textposition = 'auto',
               name = 'Normal')

trace3 = go.Bar(x = Driving_Road_Condition.road_condition[Driving_Road_Condition.DrivingStyle=='3'],
               y = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='3'],
                text = Driving_Road_Condition.Count[Driving_Road_Condition.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Vague')

data = [trace1, trace2, trace3]
layout = go.Layout(width = 1000,
#                    barmode='stack',
                   height = 600,title = 'Driving Style with Road Condition',
                  xaxis =  dict(title = 'Road Condition'),
                   yaxis =  dict(title = 'Counts',range=[0, Driving_Road_Condition.Count.max()+10]))
fig = go.Figure(data=data, layout=layout)
iplot(fig)

Observation

Most of the Drivers drive Normal way But there are few people who Inspite of Wet and Snow on the Roads Drive Aggressively

In [41]:
Driving_Lighting_Condition = train.groupby(['DrivingStyle','lighting_condition']).size().to_frame()
Driving_Lighting_Condition = Driving_Lighting_Condition.reset_index()
Driving_Lighting_Condition.columns = ['DrivingStyle','lighting_condition','Count']
In [42]:
trace1 = go.Bar(x = Driving_Lighting_Condition.lighting_condition[Driving_Lighting_Condition.DrivingStyle=='1'],
           y = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='1'],
            text = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='1'],
            textposition = 'auto',
           name = 'Aggresive')
trace2 = go.Bar(x = Driving_Lighting_Condition.lighting_condition[Driving_Lighting_Condition.DrivingStyle=='2'],
               y = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='2'],
                text = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='2'],
                textposition = 'auto',
               name = 'Normal')

trace3 = go.Bar(x = Driving_Lighting_Condition.lighting_condition[Driving_Lighting_Condition.DrivingStyle=='3'],
               y = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='3'],
                text = Driving_Lighting_Condition.Count[Driving_Lighting_Condition.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Vague')

data = [trace1, trace2, trace3]
layout = go.Layout(width = 1000,
#                    barmode='stack',
                   height = 600,title = 'Driving Style with Lighting Condition',
                  xaxis =  dict(title = 'Lighting Condition'),
                   yaxis =  dict(title = 'Counts',range=[0, Driving_Lighting_Condition.Count.max()+10]))
fig = go.Figure(data=data, layout=layout)
iplot(fig)

Observation :

At Night Time Number of people driving Aggressive are more compared to other Time of the day and also at Night people drive more Vaguely

Manipulating Date Time Column

In [43]:
date =[]
time = []
for i in train.date_time_trip:
    a = str(i).split()
    date.append(str(a[0]))
    time.append(str(a[1]))
    
train['date']=date
train['time']=time
In [44]:
day = []
month = []
year = []
hour = []
for i in train.date:
    a = str(i).split('-')
    year.append(str(a[0]))
    month.append(str(a[1]))
    day.append(str(a[2]))
    
for i in train.time:
    a = str(i).split(':')
    hour.append(str(a[0]))
    
train['day'] = day
train['month'] = month
train['year'] = year
train['hour'] = hour
In [45]:
Driving_Hour = train.groupby(['DrivingStyle','hour']).size().to_frame()
Driving_Hour = Driving_Hour.reset_index()
Driving_Hour.columns = ['DrivingStyle','Hour','Count']
In [46]:
trace1 = go.Bar(x = Driving_Hour.Hour[Driving_Hour.DrivingStyle=='1'],
           y = Driving_Hour.Count[Driving_Hour.DrivingStyle=='1'],
            text = Driving_Hour.Count[Driving_Hour.DrivingStyle=='1'],
            textposition = 'auto',
           name = 'Aggresive')
trace2 = go.Bar(x = Driving_Hour.Hour[Driving_Hour.DrivingStyle=='2'],
               y = Driving_Hour.Count[Driving_Hour.DrivingStyle=='2'],
                text = Driving_Hour.Count[Driving_Hour.DrivingStyle=='2'],
                textposition = 'auto',
               name = 'Normal')

trace3 = go.Bar(x = Driving_Hour.Hour[Driving_Hour.DrivingStyle=='3'],
               y = Driving_Hour.Count[Driving_Hour.DrivingStyle=='3'],
                text = Driving_Hour.Count[Driving_Hour.DrivingStyle=='3'],
                textposition = 'auto',
               name = 'Vague')

data = [trace1, trace2, trace3]
layout = go.Layout(width = 1000,
#                    barmode='stack',
                   height = 600,title = 'Driving Style with Hour',
                  xaxis =  dict(title = 'Hour'),
                   yaxis =  dict(title = 'Counts',range=[0, Driving_Hour.Count.max()+10]))
fig = go.Figure(data=data, layout=layout)
iplot(fig)

Observation :

There were very few people who drove between 12 pm to 6 am and comparitivele the percentage of people driving agressively was too high Between 2pm to 12am people people drive more vaguely

In [47]:
plot_bar('month')

Observation :

No data for the month of June to October

Relationship between Preceeding Vehicle Speed and Vehicle Speed

In [48]:
temp = train.sample(frac=0.01)
X = temp.vehicle_speed
Y = temp.pre_vehicle_speed
plt.figure(figsize=(15,10))
plt.xlabel("Vehicle speed")
plt.ylabel("Preceeding Vehicle speed")
plt.title("Vehicle speed vs Preceeding Vehicle speed")
plt.scatter(X,Y)
plt.show()

Observation :

The speed of preceeding Vehicle and the current vehicle are almost same

Relation between Month and road condition

In [49]:
sns.catplot(x="road_condition", y="month",data=train,);

Observation :

No Snow in the month of May and November

Trivariate Plots

Relation between Driving Style Road Condition and Vehicle Speed

In [50]:
# temp = train.sample(frac=0.01)
# sns.catplot(x="DrivingStyle", y="vehicle_speed",col="road_condition", aspect=.6,  kind="swarm", data=temp);

Observation :

  1. Most of the people drive when the road condition is Dry or Wet.
  2. Few people drive aggressively inspite of Snow on th road
In [51]:
# train.Num_Axle
In [52]:
temp = train.sample(frac=0.02)
sns.catplot(x="vehicle_speed", y="Num_Axle",col="DrivingStyle",  kind="swarm", data=temp);

Observation :

  1. The vehicles with 3 Axles drive less aggressively according to data given
  2. There are more vehicles with 2 axles

Feature Engineering

In [53]:
train.head()
Out[53]:
ID date_time_trip road_lane vehicle_speed pre_vehicle_ID pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle road_condition ... Vehicle_len Vehicle_Wt Num_Axle DrivingStyle date time day month year hour
0 DR_24526 2012-03-21 09:14:55 1 81 57227 87 16986 941 94.0 Dry ... 1265 10243 2 2 2012-03-21 09:14:55 21 03 2012 09
1 DR_24526 2012-03-21 09:15:07 1 88 57229 81 1708 551 11.0 Dry ... 1265 10243 2 2 2012-03-21 09:15:07 21 03 2012 09
2 DR_24526 2012-03-21 09:15:12 1 88 57230 88 22892 1698 4.0 Dry ... 1265 10243 2 2 2012-03-21 09:15:12 21 03 2012 09
3 DR_24526 2012-03-21 09:15:48 2 84 57228 89 1945 544 127.0 Dry ... 1265 10243 2 2 2012-03-21 09:15:48 21 03 2012 09
4 DR_24526 2012-03-21 09:15:54 1 89 57231 88 13787 1893 42.0 Dry ... 1265 10243 2 2 2012-03-21 09:15:54 21 03 2012 09

5 rows × 27 columns

In [54]:
train.columns
Out[54]:
Index(['ID', 'date_time_trip', 'road_lane', 'vehicle_speed', 'pre_vehicle_ID',
       'pre_vehicle_speed', 'pre_vehicle_wt', 'pre_vehicle_len',
       'time_gap_pre_vehicle', 'road_condition', 'air_temp',
       'precipitation_type', 'precipitation_intensity', 'rel_humidity',
       'wind_direction', 'wind_speed', 'lighting_condition', 'Vehicle_len',
       'Vehicle_Wt', 'Num_Axle', 'DrivingStyle', 'date', 'time', 'day',
       'month', 'year', 'hour'],
      dtype='object')
In [55]:
train.dtypes
Out[55]:
ID                          object
date_time_trip              object
road_lane                    int64
vehicle_speed                int64
pre_vehicle_ID               int64
pre_vehicle_speed            int64
pre_vehicle_wt               int64
pre_vehicle_len              int64
time_gap_pre_vehicle       float64
road_condition              object
air_temp                   float64
precipitation_type          object
precipitation_intensity     object
rel_humidity               float64
wind_direction             float64
wind_speed                 float64
lighting_condition          object
Vehicle_len                  int64
Vehicle_Wt                   int64
Num_Axle                     int64
DrivingStyle                object
date                        object
time                        object
day                         object
month                       object
year                        object
hour                        object
dtype: object

Dropping unnecessay columns

In [56]:
train = train.drop(['date_time_trip','date','time','day','year','pre_vehicle_ID','month','hour'],axis=1)
In [57]:
train.columns
Out[57]:
Index(['ID', 'road_lane', 'vehicle_speed', 'pre_vehicle_speed',
       'pre_vehicle_wt', 'pre_vehicle_len', 'time_gap_pre_vehicle',
       'road_condition', 'air_temp', 'precipitation_type',
       'precipitation_intensity', 'rel_humidity', 'wind_direction',
       'wind_speed', 'lighting_condition', 'Vehicle_len', 'Vehicle_Wt',
       'Num_Axle', 'DrivingStyle'],
      dtype='object')
In [58]:
train.dtypes
Out[58]:
ID                          object
road_lane                    int64
vehicle_speed                int64
pre_vehicle_speed            int64
pre_vehicle_wt               int64
pre_vehicle_len              int64
time_gap_pre_vehicle       float64
road_condition              object
air_temp                   float64
precipitation_type          object
precipitation_intensity     object
rel_humidity               float64
wind_direction             float64
wind_speed                 float64
lighting_condition          object
Vehicle_len                  int64
Vehicle_Wt                   int64
Num_Axle                     int64
DrivingStyle                object
dtype: object

Printing unique values for each column

If the unique value are less that 10 in a column assigning it as categorical variable else numerical variable

In [59]:
cat_cols =[]
num_cols =[]
for i in train.columns:
    if (len(np.unique(train[i])) < 10) & (i != 'month'):
        cat_cols.append(i)
        print("{} : {} : {} ".format(i,len(np.unique(train[i])),np.unique(train[i])))
    else:
        num_cols.append(i)

road_lane : 2 : [1 2] 
road_condition : 4 : ['Dry' 'Snow covered' 'Visible tracks' 'Wet'] 
precipitation_type : 3 : ['clear' 'rain' 'snow'] 
precipitation_intensity : 5 : ['High' 'Low' 'Moderate' 'None' 'NotKnown'] 
lighting_condition : 3 : ['daylight' 'night' 'twilight'] 
Num_Axle : 8 : [2 3 4 5 6 7 8 9] 
DrivingStyle : 3 : ['1' '2' '3']
In [60]:
ID = 'ID'
cat_cols.append(ID)

Convert Categorical variables to category

In [61]:
for i in cat_cols:
    train[i] = train[i].astype('category')
In [62]:
num_cols
Out[62]:
['ID',
 'vehicle_speed',
 'pre_vehicle_speed',
 'pre_vehicle_wt',
 'pre_vehicle_len',
 'time_gap_pre_vehicle',
 'air_temp',
 'rel_humidity',
 'wind_direction',
 'wind_speed',
 'Vehicle_len',
 'Vehicle_Wt']

Checking and Imputing NA's

In [63]:
train.isna().sum()
Out[63]:
ID                            0
road_lane                     0
vehicle_speed                 0
pre_vehicle_speed             0
pre_vehicle_wt                0
pre_vehicle_len               0
time_gap_pre_vehicle       2455
road_condition                0
air_temp                   2057
precipitation_type            0
precipitation_intensity       0
rel_humidity               2105
wind_direction             2114
wind_speed                 2464
lighting_condition            0
Vehicle_len                   0
Vehicle_Wt                    0
Num_Axle                      0
DrivingStyle                  0
dtype: int64
In [64]:
train.air_temp.fillna(train.air_temp.mean(), inplace=True)
train.wind_speed.fillna(train.wind_speed.mean(), inplace=True)
train.time_gap_pre_vehicle.fillna(train.time_gap_pre_vehicle.mean(), inplace=True)
train.wind_direction.fillna(train.wind_direction.mean(), inplace=True)
train.rel_humidity.fillna(train.rel_humidity.mean(), inplace=True)
In [65]:
np.unique(train.precipitation_intensity)
Out[65]:
array(['High', 'Low', 'Moderate', 'None', 'NotKnown'], dtype=object)

Aggregating rows

In [66]:
train_data_cat_aggregated =train[cat_cols].groupby(['ID'], sort=False).max()
train_data_cat = train_data_cat_aggregated.reset_index()
train_data_cat = train_data_cat.rename(columns = {'index':'ID'})
train_data_cat.head(4)
Out[66]:
ID road_lane road_condition precipitation_type precipitation_intensity lighting_condition Num_Axle DrivingStyle
0 DR_24526 2 Dry clear None daylight 2 2
1 DR_30052 2 Dry clear None daylight 6 2
2 DR_40928 2 Dry rain Low daylight 4 2
3 DR_66033 2 Dry clear None daylight 4 2
In [67]:
print(train_data_cat.shape)
train_data_cat.columns

(12994, 8)
Out[67]:
Index(['ID', 'road_lane', 'road_condition', 'precipitation_type',
       'precipitation_intensity', 'lighting_condition', 'Num_Axle',
       'DrivingStyle'],
      dtype='object')
In [68]:
num_cols
Out[68]:
['ID',
 'vehicle_speed',
 'pre_vehicle_speed',
 'pre_vehicle_wt',
 'pre_vehicle_len',
 'time_gap_pre_vehicle',
 'air_temp',
 'rel_humidity',
 'wind_direction',
 'wind_speed',
 'Vehicle_len',
 'Vehicle_Wt']
In [69]:
train_data_num_aggregated =train[num_cols].groupby(['ID'], sort=False).mean()
train_data_num = train_data_num_aggregated.reset_index()
train_data_num.head()
train_data_num.shape
Out[69]:
(12994, 12)
In [70]:
print(train_data_num.shape)
train_data_num.columns

(12994, 12)
Out[70]:
Index(['ID', 'vehicle_speed', 'pre_vehicle_speed', 'pre_vehicle_wt',
       'pre_vehicle_len', 'time_gap_pre_vehicle', 'air_temp', 'rel_humidity',
       'wind_direction', 'wind_speed', 'Vehicle_len', 'Vehicle_Wt'],
      dtype='object')
In [71]:
aggr_train_data = pd.merge(train_data_num,train_data_cat, how='outer', on=['ID'])

print(aggr_train_data.shape)
aggr_train_data.head()

(12994, 19)
Out[71]:
ID vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle air_temp rel_humidity wind_direction wind_speed Vehicle_len Vehicle_Wt road_lane road_condition precipitation_type precipitation_intensity lighting_condition Num_Axle DrivingStyle
0 DR_24526 87.473684 87.052632 8443.789474 965.473684 67.842105 7.000000 95.000000 125.157895 0.052632 1265.0 10243.0 2 Dry clear None daylight 2 2
1 DR_30052 82.272727 86.090909 6887.000000 920.636364 138.363636 4.636364 95.545455 157.818182 6.000000 1680.0 30871.0 2 Dry clear None daylight 6 2
2 DR_40928 85.090909 82.454545 4624.363636 773.909091 159.545455 1.545455 95.000000 64.454545 0.000000 936.0 3224.0 2 Dry rain Low daylight 4 2
3 DR_66033 82.818182 83.727273 9229.909091 1210.545455 118.454545 3.545455 95.000000 169.727273 1.363636 1503.0 3315.0 2 Dry clear None daylight 4 2
4 DR_45266 82.750000 82.916667 3058.416667 723.583333 109.500000 3.000000 95.000000 191.000000 3.000000 929.0 3022.0 2 Dry clear None daylight 4 2
In [72]:
target = aggr_train_data['DrivingStyle']
trainWithoutTarget = aggr_train_data.drop(['DrivingStyle','ID'],axis=1)
In [73]:
trainWithoutTarget.dtypes
Out[73]:
vehicle_speed              float64
pre_vehicle_speed          float64
pre_vehicle_wt             float64
pre_vehicle_len            float64
time_gap_pre_vehicle       float64
air_temp                   float64
rel_humidity               float64
wind_direction             float64
wind_speed                 float64
Vehicle_len                float64
Vehicle_Wt                 float64
road_lane                    int64
road_condition              object
precipitation_type          object
precipitation_intensity     object
lighting_condition          object
Num_Axle                     int64
dtype: object

Dummifying Categorical Vriable

In [74]:
train_dummified = pd.get_dummies(trainWithoutTarget)
In [75]:
train_dummified.dtypes
Out[75]:
vehicle_speed                       float64
pre_vehicle_speed                   float64
pre_vehicle_wt                      float64
pre_vehicle_len                     float64
time_gap_pre_vehicle                float64
air_temp                            float64
rel_humidity                        float64
wind_direction                      float64
wind_speed                          float64
Vehicle_len                         float64
Vehicle_Wt                          float64
road_lane                             int64
Num_Axle                              int64
road_condition_Dry                    uint8
road_condition_Snow covered           uint8
road_condition_Visible tracks         uint8
road_condition_Wet                    uint8
precipitation_type_clear              uint8
precipitation_type_rain               uint8
precipitation_type_snow               uint8
precipitation_intensity_High          uint8
precipitation_intensity_Low           uint8
precipitation_intensity_Moderate      uint8
precipitation_intensity_None          uint8
precipitation_intensity_NotKnown      uint8
lighting_condition_daylight           uint8
lighting_condition_night              uint8
lighting_condition_twilight           uint8
dtype: object

Test Data

Reading the test Data

In [76]:
test_data = pd.read_csv('Test.csv')
In [77]:
test_vehicletravelling_data = pd.read_csv('Test_Vehicletravellingdata.csv')
In [78]:
test_weather_data =pd.read_csv('Test_WeatherData.csv')

Renaming the Test data columns as per Attribute Description

In [79]:
test_data.columns = ['ID','Vehicle_len','Vehicle_Wt','Num_Axle']
In [80]:
test_vehicletravelling_data.columns = ['ID','date_time_trip','road_lane','vehicle_speed','pre_vehicle_ID','pre_vehicle_speed','pre_vehicle_wt','pre_vehicle_len','time_gap_pre_vehicle','road_condition']
In [81]:
test_weather_data.columns = ['ID','date_time_trip','air_temp','precipitation_type','precipitation_intensity','rel_humidity','wind_direction','wind_speed','lighting_condition']

Merging Data

  1. Merging test vehicle travelling data and test weather data on 'ID' and 'date and time' column
  2. Merging the above data with test data i.e test vehicle data on 'ID' column
In [82]:
test_vehicletravelling_weather = pd.merge(test_vehicletravelling_data, test_weather_data, how='inner', on=['ID','date_time_trip'])
In [83]:
test = pd.merge(test_vehicletravelling_weather,test_data, how='outer', on=['ID'])
In [84]:
test = test.drop(['date_time_trip','pre_vehicle_ID'],axis=1)

Manipulating Date and Time in Test data

In [85]:
# date_test =[]
# time_test = []
# for i in test.date_time_trip:
#     a = str(i).split()
#     date_test.append(str(a[0]))
#     time_test.append(str(a[1]))
# test['date']=date_test
# test['time']=time_test
In [86]:
# month_test = []
# hour_test = []
# for i in test.date:
#     a = str(i).split('-')
#     month_test.append(str(a[1]))

# for j in test.time:
#     b = str(j).split(':')
#     hour_test.append(str(b[0]))

# test['month'] = month_test
# test['hour'] = hour_test

Checking unique values for each column in Test Data

If the unique value are less that 10 in a column assigning it as categorical variable else numerical variable

In [87]:
cat_cols_test =[]
num_cols_test =[]
for i in test.columns:
    if (len(np.unique(test[i])) < 10) :
        cat_cols_test.append(i)
        print("{} : {} : {} ".format(i,len(np.unique(test[i])),np.unique(test[i])))
    else:
        num_cols_test.append(i)
        
ID = 'ID'
cat_cols_test.append(ID)

road_lane : 2 : [1 2] 
road_condition : 4 : ['Dry' 'Snow covered' 'Visible tracks' 'Wet'] 
precipitation_type : 3 : ['clear' 'rain' 'snow'] 
precipitation_intensity : 5 : [' ' 'High' 'Low' 'Moderate' 'None'] 
lighting_condition : 3 : ['daylight' 'night' 'twilight'] 
Num_Axle : 7 : [2 3 4 5 6 7 9]

Convert Categorical variables to category in Test Data

In [88]:
for i in cat_cols_test:
    test[i] = test[i].astype('category')
In [89]:
test.dtypes
Out[89]:
ID                         category
road_lane                  category
vehicle_speed                 int64
pre_vehicle_speed             int64
pre_vehicle_wt                int64
pre_vehicle_len               int64
time_gap_pre_vehicle        float64
road_condition             category
air_temp                    float64
precipitation_type         category
precipitation_intensity    category
rel_humidity                float64
wind_direction              float64
wind_speed                  float64
lighting_condition         category
Vehicle_len                   int64
Vehicle_Wt                    int64
Num_Axle                   category
dtype: object
In [90]:
# test[datetimecols] = test[datetimecols].apply(pd.to_numeric)
# test.dtypes
In [91]:
num_cols_test
Out[91]:
['ID',
 'vehicle_speed',
 'pre_vehicle_speed',
 'pre_vehicle_wt',
 'pre_vehicle_len',
 'time_gap_pre_vehicle',
 'air_temp',
 'rel_humidity',
 'wind_direction',
 'wind_speed',
 'Vehicle_len',
 'Vehicle_Wt']

Checking and Imputing NA's on Test Data

In [92]:
test.isna().sum()
Out[92]:
ID                           0
road_lane                    0
vehicle_speed                0
pre_vehicle_speed            0
pre_vehicle_wt               0
pre_vehicle_len              0
time_gap_pre_vehicle       956
road_condition               0
air_temp                   934
precipitation_type           0
precipitation_intensity      0
rel_humidity               943
wind_direction             964
wind_speed                 983
lighting_condition           0
Vehicle_len                  0
Vehicle_Wt                   0
Num_Axle                     0
dtype: int64
In [93]:
test.time_gap_pre_vehicle.fillna(test.time_gap_pre_vehicle.mean(), inplace=True)
test.air_temp.fillna(test.air_temp.mean(), inplace=True)
test.rel_humidity.fillna(test.rel_humidity.mean(), inplace=True)
test.wind_direction.fillna(test.wind_direction.mean(), inplace=True)
test.wind_speed.fillna(test.wind_speed.mean(), inplace=True)
In [94]:
np.unique(test.precipitation_intensity)
Out[94]:
array([' ', 'High', 'Low', 'Moderate', 'None'], dtype=object)
In [95]:
test.precipitation_intensity = test.precipitation_intensity.replace(to_replace=' ', value='NotKnown').astype('category')

Aggregating test data

In [96]:
test_data_cat_aggregated =test[cat_cols_test].groupby(['ID'], sort=False).max()
test_data_cat = test_data_cat_aggregated.reset_index()
test_data_cat = test_data_cat.rename(columns = {'index':'ID'})
test_data_cat.head(4)
Out[96]:
ID road_lane road_condition precipitation_type precipitation_intensity lighting_condition Num_Axle
0 DR_79986 2 Wet clear None daylight 2
1 DR_82800 2 Wet clear None daylight 2
2 DR_93526 2 Wet clear None daylight 5
3 DR_94423 2 Wet clear None daylight 2
In [97]:
test_data_cat.dtypes
Out[97]:
ID                         category
road_lane                     int64
road_condition               object
precipitation_type           object
precipitation_intensity      object
lighting_condition           object
Num_Axle                      int64
dtype: object
In [98]:
test_data_num_aggregated =test[num_cols_test].groupby(['ID'], sort=False).mean()
test_data_num = test_data_num_aggregated.reset_index()

print(test_data_num.shape)
test_data_num.head()

(4880, 12)
Out[98]:
ID vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle air_temp rel_humidity wind_direction wind_speed Vehicle_len Vehicle_Wt
0 DR_79986 88.000000 88.230769 5974.846154 999.153846 85.846154 6.615385 95.0 179.538462 6.307692 574.0 1268.0
1 DR_82800 82.625000 83.500000 4366.750000 724.875000 124.625000 4.000000 95.0 230.250000 4.000000 530.0 2379.0
2 DR_93526 84.000000 83.571429 11414.214286 1025.214286 107.500000 3.000000 95.0 231.000000 4.000000 1687.0 17648.0
3 DR_94423 84.333333 84.333333 9380.916667 1034.083333 106.583333 8.500000 94.0 193.833333 5.083333 1353.0 13525.0
4 DR_93909 87.357143 88.785714 10362.357143 957.357143 81.857143 2.000000 93.0 79.000000 6.000000 597.0 2280.0
In [99]:
test_data_num.columns
Out[99]:
Index(['ID', 'vehicle_speed', 'pre_vehicle_speed', 'pre_vehicle_wt',
       'pre_vehicle_len', 'time_gap_pre_vehicle', 'air_temp', 'rel_humidity',
       'wind_direction', 'wind_speed', 'Vehicle_len', 'Vehicle_Wt'],
      dtype='object')
In [100]:
test_data_final = pd.merge(test_data_num,test_data_cat ,on=['ID'],how='outer')
test_data_final = test_data_final.drop(['ID'],axis=1)
In [101]:
# test_data_ID.dtypes

Dummifying Categorical Vriable

In [102]:
test_dummified = pd.get_dummies(test_data_final)
In [103]:
train_dummified.dtypes
Out[103]:
vehicle_speed                       float64
pre_vehicle_speed                   float64
pre_vehicle_wt                      float64
pre_vehicle_len                     float64
time_gap_pre_vehicle                float64
air_temp                            float64
rel_humidity                        float64
wind_direction                      float64
wind_speed                          float64
Vehicle_len                         float64
Vehicle_Wt                          float64
road_lane                             int64
Num_Axle                              int64
road_condition_Dry                    uint8
road_condition_Snow covered           uint8
road_condition_Visible tracks         uint8
road_condition_Wet                    uint8
precipitation_type_clear              uint8
precipitation_type_rain               uint8
precipitation_type_snow               uint8
precipitation_intensity_High          uint8
precipitation_intensity_Low           uint8
precipitation_intensity_Moderate      uint8
precipitation_intensity_None          uint8
precipitation_intensity_NotKnown      uint8
lighting_condition_daylight           uint8
lighting_condition_night              uint8
lighting_condition_twilight           uint8
dtype: object
In [104]:
test_dummified.dtypes
Out[104]:
vehicle_speed                       float64
pre_vehicle_speed                   float64
pre_vehicle_wt                      float64
pre_vehicle_len                     float64
time_gap_pre_vehicle                float64
air_temp                            float64
rel_humidity                        float64
wind_direction                      float64
wind_speed                          float64
Vehicle_len                         float64
Vehicle_Wt                          float64
road_lane                             int64
Num_Axle                              int64
road_condition_Dry                    uint8
road_condition_Snow covered           uint8
road_condition_Visible tracks         uint8
road_condition_Wet                    uint8
precipitation_type_clear              uint8
precipitation_type_rain               uint8
precipitation_type_snow               uint8
precipitation_intensity_High          uint8
precipitation_intensity_Low           uint8
precipitation_intensity_Moderate      uint8
precipitation_intensity_None          uint8
precipitation_intensity_NotKnown      uint8
lighting_condition_daylight           uint8
lighting_condition_night              uint8
lighting_condition_twilight           uint8
dtype: object
In [105]:
test_dummified.shape
Out[105]:
(4880, 28)
In [106]:
train_dummified.shape
Out[106]:
(12994, 28)

Train dummified has an extra column Num_Axle_8 So adding the same in Test Data

In [107]:
# test_dummified['Num_Axle_8'] = 0
In [108]:
test_dummified.head(4)
Out[108]:
vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle air_temp rel_humidity wind_direction wind_speed Vehicle_len ... precipitation_type_rain precipitation_type_snow precipitation_intensity_High precipitation_intensity_Low precipitation_intensity_Moderate precipitation_intensity_None precipitation_intensity_NotKnown lighting_condition_daylight lighting_condition_night lighting_condition_twilight
0 88.000000 88.230769 5974.846154 999.153846 85.846154 6.615385 95.0 179.538462 6.307692 574.0 ... 0 0 0 0 0 1 0 1 0 0
1 82.625000 83.500000 4366.750000 724.875000 124.625000 4.000000 95.0 230.250000 4.000000 530.0 ... 0 0 0 0 0 1 0 1 0 0
2 84.000000 83.571429 11414.214286 1025.214286 107.500000 3.000000 95.0 231.000000 4.000000 1687.0 ... 0 0 0 0 0 1 0 1 0 0
3 84.333333 84.333333 9380.916667 1034.083333 106.583333 8.500000 94.0 193.833333 5.083333 1353.0 ... 0 0 0 0 0 1 0 1 0 0

4 rows × 28 columns

In [109]:
test.isna().sum()
Out[109]:
ID                         0
road_lane                  0
vehicle_speed              0
pre_vehicle_speed          0
pre_vehicle_wt             0
pre_vehicle_len            0
time_gap_pre_vehicle       0
road_condition             0
air_temp                   0
precipitation_type         0
precipitation_intensity    0
rel_humidity               0
wind_direction             0
wind_speed                 0
lighting_condition         0
Vehicle_len                0
Vehicle_Wt                 0
Num_Axle                   0
dtype: int64

All reprocessing over

Train Test Split

In [149]:
train_dummified.head()
Out[149]:
vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle air_temp rel_humidity wind_direction wind_speed Vehicle_len ... precipitation_type_rain precipitation_type_snow precipitation_intensity_High precipitation_intensity_Low precipitation_intensity_Moderate precipitation_intensity_None precipitation_intensity_NotKnown lighting_condition_daylight lighting_condition_night lighting_condition_twilight
0 87.473684 87.052632 8443.789474 965.473684 67.842105 7.000000 95.000000 125.157895 0.052632 1265.0 ... 0 0 0 0 0 1 0 1 0 0
1 82.272727 86.090909 6887.000000 920.636364 138.363636 4.636364 95.545455 157.818182 6.000000 1680.0 ... 0 0 0 0 0 1 0 1 0 0
2 85.090909 82.454545 4624.363636 773.909091 159.545455 1.545455 95.000000 64.454545 0.000000 936.0 ... 1 0 0 1 0 0 0 1 0 0
3 82.818182 83.727273 9229.909091 1210.545455 118.454545 3.545455 95.000000 169.727273 1.363636 1503.0 ... 0 0 0 0 0 1 0 1 0 0
4 82.750000 82.916667 3058.416667 723.583333 109.500000 3.000000 95.000000 191.000000 3.000000 929.0 ... 0 0 0 0 0 1 0 1 0 0

5 rows × 28 columns

In [147]:
import random

random.seed(1234) 
x_train, x_val, y_train, y_val = train_test_split(train_dummified, target, stratify=target, test_size=0.25)
In [148]:
x_train.head()
Out[148]:
vehicle_speed pre_vehicle_speed pre_vehicle_wt pre_vehicle_len time_gap_pre_vehicle air_temp rel_humidity wind_direction wind_speed Vehicle_len ... precipitation_type_rain precipitation_type_snow precipitation_intensity_High precipitation_intensity_Low precipitation_intensity_Moderate precipitation_intensity_None precipitation_intensity_NotKnown lighting_condition_daylight lighting_condition_night lighting_condition_twilight
4486 85.666667 85.916667 8202.333333 947.0 140.75 2.0 93.0 163.0 9.0 939.0 ... 0 0 0 0 0 1 0 0 0 1
7645 87.000000 93.500000 10392.500000 1268.5 340.50 7.0 55.0 191.0 3.0 1968.0 ... 0 0 0 0 0 1 0 0 1 0
9594 78.000000 59.500000 1875.000000 710.0 208.00 4.0 91.0 242.0 1.0 521.0 ... 0 1 0 1 0 0 0 0 1 0
6033 93.000000 92.600000 6797.600000 979.2 203.40 4.8 43.2 191.6 5.8 553.0 ... 0 0 0 0 0 1 0 0 1 0
9393 84.500000 98.500000 13100.000000 1176.0 266.00 2.0 46.0 197.0 6.5 1942.0 ... 0 0 0 0 0 1 0 0 1 0

5 rows × 28 columns

MODEL BUILDING

Logistic Regression

In [111]:
logistic_model = LogisticRegression(random_state=0, solver='newton-cg', multi_class='multinomial')
logistic_model.fit(x_train, y_train)
# logistic_model.

C:\Users\ADMIN\Anaconda3\lib\site-packages\sklearn\utils\optimize.py:203: ConvergenceWarning:

newton-cg failed to converge. Increase the number of iterations.
Out[111]:
LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
          intercept_scaling=1, max_iter=100, multi_class='multinomial',
          n_jobs=1, penalty='l2', random_state=0, solver='newton-cg',
          tol=0.0001, verbose=0, warm_start=False)
In [112]:
predictions_log_model_train = logistic_model.predict(x_train)
predictions_log_model_val = logistic_model.predict(x_val)

print(recall_score(y_train,predictions_log_model_train,labels=[1],average='macro'))
print(recall_score(y_val,predictions_log_model_val,labels=[1],average='macro'))

0.3786641037962518
0.354978354978355

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
In [113]:
#### Decision Tree
In [164]:
from sklearn.model_selection import GridSearchCV
parameters = {'max_depth':[7],'min_samples_split' : [10],'criterion':['gini']}
clf = GridSearchCV(DecisionTreeClassifier(), parameters, n_jobs=4,cv=5)
clf.fit(X=x_train, y=y_train)
tree_model = clf.best_estimator_
print (clf.best_score_, clf.best_params_) 

0.719138019497178 {'criterion': 'gini', 'max_depth': 7, 'min_samples_split': 10}
In [165]:
predictions_tree_model_train = tree_model.predict(x_train)
predictions_tree_model_val = tree_model.predict(x_val)

print(recall_score(y_train,predictions_tree_model_train,labels=[1],average='macro'))
print(recall_score(y_val,predictions_tree_model_val,labels=[1],average='macro'))

0.561268620855358
0.47763347763347763

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison

Random Forest

In [116]:
rf = RandomForestClassifier(n_estimators = 1500, n_jobs=-1,random_state = 42)
rf.fit(x_train, y_train)
Out[116]:
RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
            max_depth=None, max_features='auto', max_leaf_nodes=None,
            min_impurity_decrease=0.0, min_impurity_split=None,
            min_samples_leaf=1, min_samples_split=2,
            min_weight_fraction_leaf=0.0, n_estimators=1500, n_jobs=-1,
            oob_score=False, random_state=42, verbose=0, warm_start=False)
In [117]:
predictions_rf_train = rf.predict(x_train)
predictions_rf_val = rf.predict(x_val)

print(recall_score(y_train,predictions_rf_train,labels=[1],average='macro'))
print(recall_score(y_val,predictions_rf_val,labels=[1],average='macro'))

1.0
0.5108225108225108

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
In [118]:
from sklearn.model_selection import RandomizedSearchCV
from pprint import pprint
from sklearn.ensemble import RandomForestClassifier
n_estimators = [int(x) for x in np.linspace(start = 200, stop = 2000, num = 10)]
max_features = ['auto', 'sqrt']
max_depth = [int(x) for x in np.linspace(10, 110, num = 11)]
max_depth.append(None)
min_samples_split = [2, 5, 10]
min_samples_leaf = [1, 2, 4]
bootstrap = [True, False]
random_grid = {'n_estimators': n_estimators,
               'max_features': max_features,
               'max_depth': max_depth,
               'min_samples_split': min_samples_split,
               'min_samples_leaf': min_samples_leaf,
               'bootstrap': bootstrap}
pprint(random_grid)

{'bootstrap': [True, False],
 'max_depth': [10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, None],
 'max_features': ['auto', 'sqrt'],
 'min_samples_leaf': [1, 2, 4],
 'min_samples_split': [2, 5, 10],
 'n_estimators': [200, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000]}
In [121]:
rf = RandomForestClassifier(random_state = 42)
rf_random = RandomizedSearchCV(estimator = rf, param_distributions = random_grid, n_iter = 1, cv = 3, verbose=2, random_state=42)
In [123]:
rf_random.fit(x_train, y_train)
best_rf1= rf_random.best_estimator_
best_rf_features = rf_random.best_params_

Fitting 3 folds for each of 1 candidates, totalling 3 fits
[CV] n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True 
[CV]  n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True, total=   3.7s
[CV] n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True 

[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    4.1s remaining:    0.0s

[CV]  n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True, total=   3.7s
[CV] n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True 
[CV]  n_estimators=200, min_samples_split=10, min_samples_leaf=2, max_features=sqrt, max_depth=50, bootstrap=True, total=   3.7s

[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:   12.5s finished
In [124]:
predictions_bestrf_train1 = best_rf1.predict(x_train)
predictions_bestrf_val1 = best_rf1.predict(x_val)

print(recall_score(y_train,predictions_bestrf_train1,labels=[1],average='macro'))
print(recall_score(y_val,predictions_bestrf_val1,labels=[1],average='macro'))

0.8077847188851514
0.5108225108225108

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
In [130]:
predictions_bestrf_tst = best_rf1.predict(test_dummified)
In [131]:
test_data_ID = test_data_num.ID
In [132]:
rf_grid2= pd.DataFrame({'ID':test_data_ID})
rf_grid2['DrivingStyle'] = predictions_bestrf_tst
In [ ]:
#pd.DataFrame(rf_grid3).to_csv("anita_rf_grid12.csv")

Gradient Boosting

In [133]:
learning_rates = [0.05, 0.1, 0.25, 0.5]
for learning_rate in learning_rates:
    gb = GradientBoostingClassifier(n_estimators=125, learning_rate = learning_rate,min_samples_split=4, max_features=None, max_depth = 3, random_state = 20)
    gb.fit(x_train, y_train)
    predictions_gb_train = gb.predict(x_train)
    predictions_gb_val = gb.predict(x_val)

    print(recall_score(y_train,predictions_gb_train,labels=[1],average='macro'))
    print(recall_score(y_val,predictions_gb_val,labels=[1],average='macro'))

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison


0.565112926477655
0.49927849927849927
0.5920230658337338
0.4963924963924964
0.6665064872657376
0.4805194805194805
0.7611725132148006
0.4588744588744589
In [134]:
xgb_model = GradientBoostingClassifier(n_estimators=125, learning_rate = 0.1,min_samples_split=4, max_features=None, max_depth = 3, random_state = 20)
kfold = StratifiedKFold(n_splits=5, random_state=7)
results = cross_val_score(xgb_model, x_train, y_train, cv=kfold)
In [135]:
predictions_gb_train = gb.predict(x_train)
predictions_gb_val = gb.predict(x_val)

print(recall_score(y_train,predictions_gb_train,labels=[1],average='macro'))
print(recall_score(y_val,predictions_gb_val,labels=[1],average='macro'))

0.7611725132148006
0.4588744588744589

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
In [136]:
#### XGB
In [137]:
xgbm = XGBClassifier(max_depth=3, n_estimators=300, learning_rate=0.05).fit(x_train, y_train)
pr_train = xgbm.predict(x_train)
pr_val = xgbm.predict(x_val)

print(recall_score(y_train,pr_train,labels=[1],average='macro'))
print(recall_score(y_val,pr_val,labels=[1],average='macro'))

C:\Users\ADMIN\Anaconda3\lib\site-packages\sklearn\preprocessing\label.py:151: DeprecationWarning:

The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.

C:\Users\ADMIN\Anaconda3\lib\site-packages\sklearn\preprocessing\label.py:151: DeprecationWarning:

The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.


0.5780874579529073
0.49783549783549785

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison

Use of corss validation

Used in all the above model but it was not useful in increasing the recall rate

MLP Classifier

In [153]:
mlp = MLPClassifier(activation='relu', alpha=0.0001, batch_size='auto', beta_1=0.9,
           beta_2=0.999, early_stopping=False, epsilon=1e-08,
           hidden_layer_sizes=(100,), learning_rate='constant',
           learning_rate_init=0.001, max_iter=500, momentum=0.9,
           nesterovs_momentum=True, power_t=0.5, random_state=50,
           shuffle=True, solver='adam', tol=0.0001, validation_fraction=0.1,
           verbose=False, warm_start=True)
mlp.fit(x_train, y_train)
mlp_pred_train = mlp.predict(x_train)
rec_score_train = recall_score(y_train,mlp_pred_train,labels=[1],average='macro')
mlp_pred_val = mlp.predict(x_val)
rec_score_test = recall_score(y_val, mlp_pred_val, labels=[1],average='macro')
print(rec_score_train)
print(rec_score_test)

0.8606439211917347
0.8441558441558441

C:\Users\ADMIN\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py:468: FutureWarning:

elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
In [ ]:
# mlp = MLPClassifier(random_state=12)
# mlp.fit(x_train, y_train)
# mlp_pred_train = mlp.predict(x_train)
# rec_score_train = recall_score(y_train,mlp_pred_train,labels=[1],average='macro')
# print(434)
# print(rec_score_train)
# mlp_pred_val = mlp.predict(x_val)
# rec_score_test = recall_score(y_val, mlp_pred_val, labels=[1],average='macro')
# print(rec_score_test)

Building Tree Framework

In [154]:
def tree_to_pseudo(tree, feature_names):

    left = tree.tree_.children_left
    right = tree.tree_.children_right
    threshold = tree.tree_.threshold
    features = [feature_names[i] for i in tree.tree_.feature]
    value = tree.tree_.value

    def recurse(left, right, threshold, features, node, depth=0):
        indent = "  " * depth
        if (threshold[node] != -2):
            print(indent,"if ( " + features[node] + " <= " + str(threshold[node]) + " ) {")
            if left[node] != -1:
                recurse (left, right, threshold, features, left[node], depth+1)
                print(indent,"} else {")
                if right[node] != -1:
                    recurse (left, right, threshold, features, right[node], depth+1)
                print(indent,"}")
        else:
            print(indent,"return " + str(value[node]))

    recurse(left, right, threshold, features, 0)
In [167]:
tree_to_pseudo(tree_model,train_dummified.columns)

 if ( lighting_condition_daylight <= 0.5 ) {
   if ( vehicle_speed <= 81.25657653808594 ) {
     if ( Vehicle_len <= 1694.5 ) {
       if ( road_condition_Dry <= 0.5 ) {
         if ( road_condition_Visible tracks <= 0.5 ) {
           if ( Vehicle_Wt <= 8244.0 ) {
             if ( vehicle_speed <= 81.22335815429688 ) {
               return [[118. 316.  18.]]
             } else {
               return [[3. 0. 0.]]
             }
           } else {
             if ( wind_direction <= 175.16665649414062 ) {
               return [[13. 18.  2.]]
             } else {
               return [[38.  8.  4.]]
             }
           }
         } else {
           if ( air_temp <= 5.027777671813965 ) {
             if ( pre_vehicle_len <= 1344.5 ) {
               return [[10. 87. 10.]]
             } else {
               return [[0. 0. 2.]]
             }
           } else {
             if ( Vehicle_Wt <= 2274.0 ) {
               return [[ 0. 41.  0.]]
             } else {
               return [[ 2. 19.  0.]]
             }
           }
         }
       } else {
         if ( pre_vehicle_speed <= 100.75 ) {
           if ( rel_humidity <= 29.330883026123047 ) {
             if ( rel_humidity <= 28.761905670166016 ) {
               return [[2. 8. 0.]]
             } else {
               return [[2. 0. 0.]]
             }
           } else {
             if ( vehicle_speed <= 81.22431182861328 ) {
               return [[ 49. 579.   3.]]
             } else {
               return [[3. 5. 0.]]
             }
           }
         } else {
           return [[2. 1. 0.]]
         }
       }
     } else {
       if ( wind_direction <= 192.5454559326172 ) {
         if ( vehicle_speed <= 74.92308044433594 ) {
           if ( rel_humidity <= 88.97058868408203 ) {
             if ( time_gap_pre_vehicle <= 56.287879943847656 ) {
               return [[0. 1. 1.]]
             } else {
               return [[41.  7.  0.]]
             }
           } else {
             if ( wind_speed <= 3.8333334922790527 ) {
               return [[ 3. 10.  3.]]
             } else {
               return [[5. 0. 0.]]
             }
           }
         } else {
           if ( Vehicle_len <= 2064.5 ) {
             if ( pre_vehicle_len <= 1021.737060546875 ) {
               return [[62. 32.  4.]]
             } else {
               return [[106. 114.   7.]]
             }
           } else {
             return [[9. 0. 0.]]
           }
         }
       } else {
         if ( road_condition_Wet <= 0.5 ) {
           if ( air_temp <= 7.845238208770752 ) {
             if ( road_lane <= 1.5 ) {
               return [[7. 9. 5.]]
             } else {
               return [[43. 23.  0.]]
             }
           } else {
             if ( vehicle_speed <= 80.46666717529297 ) {
               return [[17.  1.  0.]]
             } else {
               return [[0. 1. 0.]]
             }
           }
         } else {
           if ( Vehicle_Wt <= 5723.0 ) {
             return [[0. 1. 0.]]
           } else {
             if ( Vehicle_len <= 1711.5 ) {
               return [[9. 0. 3.]]
             } else {
               return [[56.  0.  0.]]
             }
           }
         }
       }
     }
   } else {
     if ( Vehicle_len <= 1694.5 ) {
       if ( road_condition_Wet <= 0.5 ) {
         if ( wind_speed <= 4.372685432434082 ) {
           if ( rel_humidity <= 62.477272033691406 ) {
             if ( pre_vehicle_speed <= 86.93541717529297 ) {
               return [[ 40. 460. 164.]]
             } else {
               return [[ 10.  64. 206.]]
             }
           } else {
             if ( pre_vehicle_speed <= 88.04545593261719 ) {
               return [[  8.  68. 356.]]
             } else {
               return [[ 2. 32. 69.]]
             }
           }
         } else {
           if ( wind_speed <= 5.453571319580078 ) {
             if ( rel_humidity <= 43.46428680419922 ) {
               return [[ 1. 33. 53.]]
             } else {
               return [[  8.  55. 263.]]
             }
           } else {
             if ( pre_vehicle_speed <= 108.25 ) {
               return [[  7. 166. 907.]]
             } else {
               return [[0. 3. 1.]]
             }
           }
         }
       } else {
         if ( vehicle_speed <= 86.54248046875 ) {
           if ( Vehicle_Wt <= 8036.5 ) {
             if ( rel_humidity <= 52.75 ) {
               return [[ 3. 12. 13.]]
             } else {
               return [[126.  59. 273.]]
             }
           } else {
             if ( wind_direction <= 172.33334350585938 ) {
               return [[10.  7. 18.]]
             } else {
               return [[37.  6.  9.]]
             }
           }
         } else {
           if ( lighting_condition_night <= 0.5 ) {
             if ( wind_speed <= 3.3392858505249023 ) {
               return [[ 0.  5. 14.]]
             } else {
               return [[12.  3.  9.]]
             }
           } else {
             if ( Vehicle_Wt <= 15427.0 ) {
               return [[147.   0.   8.]]
             } else {
               return [[3. 0. 2.]]
             }
           }
         }
       }
     } else {
       if ( road_condition_Wet <= 0.5 ) {
         if ( vehicle_speed <= 89.53571319580078 ) {
           if ( pre_vehicle_wt <= 15813.833984375 ) {
             if ( Vehicle_len <= 1946.5 ) {
               return [[262. 189. 205.]]
             } else {
               return [[110. 120.  57.]]
             }
           } else {
             if ( rel_humidity <= 48.375 ) {
               return [[20. 22. 12.]]
             } else {
               return [[22. 68. 19.]]
             }
           }
         } else {
           if ( time_gap_pre_vehicle <= 753.25 ) {
             if ( Vehicle_Wt <= 10806.0 ) {
               return [[0. 1. 5.]]
             } else {
               return [[76. 16. 51.]]
             }
           } else {
             if ( vehicle_speed <= 98.33332824707031 ) {
               return [[13.  1.  0.]]
             } else {
               return [[0. 1. 0.]]
             }
           }
         }
       } else {
         if ( wind_direction <= 168.75 ) {
           if ( vehicle_speed <= 86.56817626953125 ) {
             if ( time_gap_pre_vehicle <= 139.66665649414062 ) {
               return [[ 9.  6. 21.]]
             } else {
               return [[27. 20.  8.]]
             }
           } else {
             if ( pre_vehicle_speed <= 92.22618865966797 ) {
               return [[23.  0.  1.]]
             } else {
               return [[1. 0. 2.]]
             }
           }
         } else {
           if ( vehicle_speed <= 81.67707824707031 ) {
             return [[2. 3. 3.]]
           } else {
             if ( lighting_condition_night <= 0.5 ) {
               return [[7. 3. 3.]]
             } else {
               return [[148.   4.   9.]]
             }
           }
         }
       }
     }
   }
 } else {
   if ( Vehicle_len <= 1694.5 ) {
     if ( road_condition_Visible tracks <= 0.5 ) {
       if ( pre_vehicle_len <= 447.16668701171875 ) {
         return [[1. 0. 0.]]
       } else {
         if ( Vehicle_Wt <= 50833.0 ) {
           if ( rel_humidity <= 18.77777862548828 ) {
             return [[1. 0. 0.]]
           } else {
             if ( vehicle_speed <= 101.3375015258789 ) {
               return [[ 186. 1877.    2.]]
             } else {
               return [[1. 0. 0.]]
             }
           }
         } else {
           return [[1. 0. 0.]]
         }
       }
     } else {
       if ( vehicle_speed <= 81.26144409179688 ) {
         if ( time_gap_pre_vehicle <= 196.08929443359375 ) {
           return [[ 0. 51.  0.]]
         } else {
           if ( time_gap_pre_vehicle <= 203.875 ) {
             return [[1. 0. 0.]]
           } else {
             if ( wind_direction <= 47.375 ) {
               return [[1. 1. 0.]]
             } else {
               return [[ 0. 10.  0.]]
             }
           }
         }
       } else {
         if ( air_temp <= 7.1666669845581055 ) {
           if ( rel_humidity <= 46.1726188659668 ) {
             return [[1. 0. 0.]]
           } else {
             return [[ 0.  0. 14.]]
           }
         } else {
           return [[0. 2. 1.]]
         }
       }
     }
   } else {
     if ( pre_vehicle_speed <= 80.87301635742188 ) {
       if ( wind_speed <= 1.7207791805267334 ) {
         if ( pre_vehicle_wt <= 6985.16650390625 ) {
           return [[3. 1. 0.]]
         } else {
           return [[0. 8. 0.]]
         }
       } else {
         if ( pre_vehicle_wt <= 17187.375 ) {
           if ( pre_vehicle_len <= 944.4880981445312 ) {
             if ( time_gap_pre_vehicle <= 126.79669189453125 ) {
               return [[5. 0. 0.]]
             } else {
               return [[ 6. 10.  0.]]
             }
           } else {
             if ( road_condition_Snow covered <= 0.5 ) {
               return [[42.  7.  1.]]
             } else {
               return [[0. 2. 0.]]
             }
           }
         } else {
           return [[0. 3. 0.]]
         }
       }
     } else {
       if ( pre_vehicle_speed <= 87.09091186523438 ) {
         if ( vehicle_speed <= 83.57777404785156 ) {
           if ( pre_vehicle_speed <= 81.400634765625 ) {
             if ( wind_speed <= 0.7272727489471436 ) {
               return [[2. 0. 0.]]
             } else {
               return [[ 0. 11.  0.]]
             }
           } else {
             if ( time_gap_pre_vehicle <= 172.375 ) {
               return [[41. 25.  1.]]
             } else {
               return [[ 4. 12.  0.]]
             }
           }
         } else {
           if ( time_gap_pre_vehicle <= 72.06128692626953 ) {
             return [[3. 0. 0.]]
           } else {
             if ( Vehicle_len <= 2040.0 ) {
               return [[24. 78.  0.]]
             } else {
               return [[0. 0. 2.]]
             }
           }
         }
       } else {
         if ( Vehicle_len <= 1706.0 ) {
           return [[0. 4. 1.]]
         } else {
           if ( air_temp <= 6.733333587646484 ) {
             if ( rel_humidity <= 93.0 ) {
               return [[23.  3.  1.]]
             } else {
               return [[0. 0. 1.]]
             }
           } else {
             return [[1. 3. 1.]]
           }
         }
       }
     }
   }
 }

Observation:

    • lighting_condition_daylight <= 0.5
    • vehicle_speed <= 81.25657653808594)
    • Vehicle_len <= 1694.5 )
    • road_condition_Dry <= 0.5 )
    • road_condition_Visible tracks <= 0.5 )
    • Vehicle_Wt <= 8244.0
    • vehicle_speed > 81.22335815429688
    • lighting_condition_daylight <= 0.5
    • vehicle_speed <= 81.25657653808594
    • Vehicle_len <= 1694.5
    • road_condition_Dry <= 0.5
    • road_condition_Visible tracks <= 0.5
    • Vehicle_Wt > 8244.0
    • wind_direction <= 175.16665649414062
    • lighting_condition_daylight <= 0.5
    • vehicle_speed <= 81.25657653808594
    • Vehicle_len <= 1694.5
    • road_condition_Dry > 0.5
    • pre_vehicle_speed <= 100.75
    • rel_humidity <= 29.330883026123047
    • rel_humidity <= 28.761905670166016

4.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len <= 1694.5
  • road_condition_Dry > 0.5
  • pre_vehicle_speed > 100.75
    • lighting_condition_daylight <= 0.5
    • vehicle_speed <= 81.25657653808594
    • Vehicle_len > 1694.5
    • wind_direction <= 192.5454559326172
    • vehicle_speed <= 74.92308044433594
    • rel_humidity <= 88.97058868408203
    • time_gap_pre_vehicle > 56.287879943847656

6.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len > 1694.5
  • wind_direction <= 192.5454559326172
  • vehicle_speed <= 74.92308044433594
  • rel_humidity > 88.97058868408203
  • wind_speed > 3.8333334922790527

7.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len > 1694.5
  • wind_direction <= 192.5454559326172
  • vehicle_speed > 74.92308044433594
  • Vehicle_len <= 2064.5
  • pre_vehicle_len <= 1021.737060546875

8.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len > 1694.5
  • wind_direction <= 192.5454559326172
  • vehicle_speed > 74.92308044433594
  • Vehicle_len > 2064.5 ) then the Driving style is Aggressive

9.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len > 1694.5
  • wind_direction > 192.5454559326172
  • road_condition_Wet <= 0.5
  • air_temp <= 7.845238208770752
  • road_lane = 2

10.

  • lighting_condition_daylight <= 0.5
  • vehicle_speed <= 81.25657653808594
  • Vehicle_len > 1694.5
  • wind_direction > 192.5454559326172
  • road_condition_Wet <= 0.5
  • Vehicle_Wt <= 5723.0
  • Vehicle_len <= 1711.5
In [156]:
import os     
os.environ["PATH"] += os.pathsep + 'C:/Users/ADMIN/Anaconda3/Library/bin/graphviz'
In [160]:
train_dummified.columns
Out[160]:
Index(['vehicle_speed', 'pre_vehicle_speed', 'pre_vehicle_wt',
       'pre_vehicle_len', 'time_gap_pre_vehicle', 'air_temp', 'rel_humidity',
       'wind_direction', 'wind_speed', 'Vehicle_len', 'Vehicle_Wt',
       'road_lane', 'Num_Axle', 'road_condition_Dry',
       'road_condition_Snow covered', 'road_condition_Visible tracks',
       'road_condition_Wet', 'precipitation_type_clear',
       'precipitation_type_rain', 'precipitation_type_snow',
       'precipitation_intensity_High', 'precipitation_intensity_Low',
       'precipitation_intensity_Moderate', 'precipitation_intensity_None',
       'precipitation_intensity_NotKnown', 'lighting_condition_daylight',
       'lighting_condition_night', 'lighting_condition_twilight'],
      dtype='object')
In [166]:
from sklearn.externals.six import StringIO  
from IPython.display import Image  
from sklearn.tree import export_graphviz
import pydotplus
dot_data = StringIO()
export_graphviz(tree_model, out_file=dot_data, filled=True, rounded=True, special_characters=True,class_names=['Aggressive', 'Normal', 'Vague'],rotate=False,feature_names=train_dummified.columns)
graph = pydotplus.graph_from_dot_data(dot_data.getvalue())  
Image(graph.create_png())
Out[166]:

Summary and Conclusion

Neural Network model performed better with 96% recall on test data Decision tree gave 70 % Recall on test data Random forest gave 80 % on train and validation data but gave 53% on test data